Ready Player Me inspired posed 3D characters

Stable Diffusion v1.5, ControlNet v1.1, Cog model

• Cog packages machine learning models as standard containers.

This model generates 3D renders of a character from two angles. It uses ControlNet for keeping the pose consistency. It uses pre-trained model for producing Ready Player Me inspired characters with consistent face and back renders.

First, download the pre-trained weights:

cog run script/download-weights 

Then, you can run predictions:

Example 1

cog predict \
-i prompt="rebel-biker" \
-i body_type="masculine" \
-i negative_prompt="handbag"

Example 2

cog predict \
-i prompt="cyberpunk goth model" \
-i body_type="feminine" \
-i negative_prompt="high contrast, dark shadows"

Features

1. Long prompts
2. Emphasizing words with round brackets and lowering the impact of the word with square brackets (Automatic1111 style)